Book Summary: Good Strategy Bad Strategy

Book: Good Strategy Bad Strategy

Author: Richard Rumelt

Key takeaways: I picked up this book not thinking much of it, but was very surprised by how much I learned, which actually is a testament to the pervasiveness of bad strategy in the corporate world (something that Rumelt talks about right away). He talks about what is a good strategy, which actually would not seem very interesting to someone unless they have first read about what bad strategy looks like, and why it is so common.

Good and Bad Strategy

I found this to be the best part of the book. Rumelt talks about how many equate strategy with broad goals that don’t mean much and merely pay homage to idea of ambition, mission and values, or much worse, simply success. Slogans like “Let’s win” and “Be #1 supplier in the industry” is not a strategy. Goal setting is not strategy.

The reasons that we have so much bad strategy is because of three reasons–(a)pain/difficulty of choice; saying no is difficult psychological, political and organizational work, (b)ease of template-style strategy, and (c) the belief that all you need to succeed is a positive mental attitude.

Rumelt says that first advantage of a good strategy is that it is unexpected; instead organizations often have broad goals and no approach but “spend more and try harder”. Good strategy should mean a cohesive response to an important challenge, and should include a set of coherent actions–these are not “implementation” details, but the punch in the strategy. Kernel of a strategy includes a diagnosis of the challenge(s), a guiding policy (which, like guardrails on highway, directs and constrains action without defining its content) and coherent action.

A guiding policy creates advantage by anticipating others’ reactions, reducing complexity and ambiguity in the situation, and exploits leverage inherent in focusing actions, when natural workings of a system, almost by design, goes scattershot. Decentralized decision making can fail when cost and benefit of actions are not borne by decentralized actors. But, this is not to say that centralized decision making is always the “best way”; coordination is costly because it fights against the gains to specialization, the most basic economies in an organized activity, and hence it needs to be pursued when gains are very high.

Unsurprisingly, good strategy requires good leaders. If a leader lists underperformance as a challenge, then the exercise is doomed because underperformance is the result of the true challenges. Good strategy requires leaders who are willing and able to say no to wide variety of actions and interests.

Sources of Strength

A strategy doesn’t just draw on existing strength; it creates strength from coherence of its design–very few companies actually do this. It also creates strength from subtle shifts in viewpoint that can reparametrize new patterns of advantage and weakness (e.g. how Wal-Mart changed the paradigm from “you need population of 100k to support a store” to thinking about networks of 150 stores that operated and negotiated as a common unit and could serve millions, plus had many things like bar-code scanners and inventory management innovation that all worked together to further its strategy.) A great strategy is often very hard for competitors to follow as it is very hard to change one’s doctrine absent near-death experience, which is why other retailers languished while Wal-Mart charged ahead.

Apart from leverage, a good strategy extracts strengths by doing the below –

  • sets proximate objectives: landing on moon, for example
  • resolves ambiguity even when the “right” answer is not known, because “engineers can’t work without a specification”, which holds true for all organized human effort
  • sets a well-oiled chain-link system, which is very hard to replicate (e.g. how Ikea has grown continually) because copying just one part doesn’t do any good
  • it designs; strategy is often thought of as a decision making process, as if selecting from a list of available options, but often it means devising a novel response from ground up (e.g. Hannibal’s military strategy against the Romans).

One interesting example that Rumelt talks about is how the U.S. put tremendous pressure on the U.S.S.R. by investing in new technology where it had an advantage and forced U.S.S.R. to invest and spend in area where they did not have a natural advantage. This was an idea expressed by Andy Marshall and James Roche of the DoD, who led the strategy of using relative advantage to impose out-of-proportion costs on the opposition.

A good strategy creates value by making the following changes –

  • Deepening advantages: Obvious, but two main reasons that this process stalls is because management can mistakenly believe that improvement can be accomplished by pressure or incentives alone. Rumelt talks about how bricklaying was done the same way for thousands of years until 1909 when it was made 2x more efficient by Gilberth; it had to re-examined from bottom up to get the improvement. Second reason is that if change is easy for others to copy then it is of little use; improvements must be protected or embedded.
  • Broadening the extent of the advantage: again, obvious–new fields and new competitions.
  • Creating higher demand
  • Strengthening isolating mechanisms: again, making it harder for competition to use your competitive advantage

Rumelt has a very interesting section which talks about how a company can navigate a wave successfully. He says clearly that it is very difficult to work with industry-wide or economy-wide changes, but the successful leaders don’t need to get it all right, but only more right than the rivals. He mentions the following guideposts –

  • Escalating fixed costs: forces the industry to consolidate because only the largest players can cover the fixed charges (something like this is underway in the semiconductor industry where a single fab can now cost $5bn+)
  • Deregulation: highly regulated companies do not know their own cost and usually develop complex systems to justify whatever they have, and it can take years for a regulated company or a former monopolist to shed excess costs.
  • Predictable biases in forecasting: the usual–continuation of the “normal”, battle of the titans where it is assumed that the largest players will duke it out, and where everyone will follow what the biggest/most profitable players is doing.
  • Incumbent response: firms will work hard to avoid changes that will undermine the complex skills that it has developed and accumulated over time.
  • Attractor state: this describes how the industry should work in the ideal state, and though it doesn’t mean that this state will come to being but it does reflect a gravity-like pull. A company that brings the industry closer to this will naturally benefit.

Rumelt describes how companies without strategy can embody both inertia (in its processes) and entropy (unfocused product line, low price to please sales and long shipping schedules to please the factory).

 

Thinking Like a Strategist

Strategy cannot be treated as an exercise in deduction, as that assumes that everything worth knowing is already know and only computation is required. To develop great strategy, one needs to lean on induction, analogy, judgement and insight.

One also needs to remember that it is very easy to hang on to the first idea that comes when facing a complex situation, and even though it is painful to let go of that judgement and re-evaluate, it should be done to make sure that the very best choice is being made.

Advertisements

2 Comments

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google+ photo

You are commenting using your Google+ account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

w

Connecting to %s

This site uses Akismet to reduce spam. Learn how your comment data is processed.